Currently, the time limit of AlevinQC is hard coded to 120h here since #196 was solved via #197. In my case, this led to Slurm PartitionTimeLimit problems (basically nothing was executed because the time limit of the queue is lower than 120h). Usually one can solve this by setting the max_time parameter. However, this parameter does not affect the AlevinQC time limit.
As an experienced nextflow user, one can easily find a temporary fix through adding this to nextflow.config:
process {
withName: 'ALEVINQC' {
time = '5.h'
}
}
But I think figuring this out might be hard for users with less experience.
Some ideas for a permanent fix:
Use the check_max function for setting the limit
Assign one of the standard process labels (maybe process_low + process_long) and tell users to perform a manual override of the time limit if they really hit the time limit
I can also open a PR myself, but maybe I am missing something here, so let's discuss first
Description of the bug
Currently, the time limit of AlevinQC is hard coded to 120h here since #196 was solved via #197. In my case, this led to Slurm PartitionTimeLimit problems (basically nothing was executed because the time limit of the queue is lower than 120h). Usually one can solve this by setting the
max_time
parameter. However, this parameter does not affect the AlevinQC time limit.As an experienced nextflow user, one can easily find a temporary fix through adding this to
nextflow.config
:But I think figuring this out might be hard for users with less experience.
Some ideas for a permanent fix:
check_max
function for setting the limitprocess_low
+process_long
) and tell users to perform a manual override of the time limit if they really hit the time limitI can also open a PR myself, but maybe I am missing something here, so let's discuss first
Command used and terminal output
No response
Relevant files
No response
System information
Slurm
nf-core/scrnaseq
version 2.6.0